Correlations

library(ggplot2)
library(tidyverse)
library(GGally)
library(ggpubr)

#load the dataframes for the figure

remove cells in the edges : pos.x > 45 & < 1150 pos.y > 45 & < 1000 ##add the replicate info

aic.df <- aic.df %>% 
  mutate(exp.field = paste0(str_split(cell.id, "_", simplify = T)[,2],"_",
                            str_split(cell.id, "_", simplify = T)[,3]),
    colony = case_when(exp.field %in% c("20min_s3", "20min_s4") ~ "Replicate 1", 
                            exp.field %in% c("20min_s5", "20min_s6") ~ "Replicate 2",
                            exp.field %in% c("20min_s7", "20min_s8" , "20min_s9") ~ "Replicate 3"))

aic.df <- aic.df %>% 
  mutate(colony = ifelse(degron == "stable" & red == "pup1-rfp", 
                         case_when(exp.field %in% c("20min_s4", "20min_s5") ~ "Replicate 1",
                                   exp.field %in% c("20min_s6", "20min_s7") ~ "Replicate 2",
                                   exp.field %in% c("20min_s8", "20min_s9") ~ "Replicate 3"), colony)) 
  # filter(value < 0.05)
  

##Cellular attributes

pup1.cell.attr <- read_csv("~/plots/all_data/all_pup1_cell_attr.csv")

Rows: 19396 Columns: 29
── Column specification ───────────────────────────────────────────────────────────────────────────────────────────────────────────────────
Delimiter: ","
chr  (4): cell.id, degron, red, treatment
dbl (25): gfp.mean.bg.af.sub.new, gfp.sum.bg.af.sub, area, area.puncta, BB_B, BB_C, Elip_B, Elip_C, no.of.voxels, pos.x, pos.y, spherac...

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

###pup1-RFP background

pup1.cell.attr <- pup1.cell.attr %>% 
  mutate(exp.field = paste0(str_split(cell.id, "_", simplify = T)[,2],"_",
                            str_split(cell.id, "_", simplify = T)[,3]),
    colony = case_when(exp.field %in% c("20min_s3", "20min_s4") ~ "Replicate 1", 
                            exp.field %in% c("20min_s5", "20min_s6") ~ "Replicate 2",
                            exp.field %in% c("20min_s7", "20min_s8" , "20min_s9") ~ "Replicate 3"))

pup1.cell.attr <- pup1.cell.attr %>% 
  mutate(colony = ifelse(degron == "stable" & red == "pup1-rfp", 
                         case_when(exp.field %in% c("20min_s4", "20min_s5") ~ "Replicate 1",
                                   exp.field %in% c("20min_s6", "20min_s7") ~ "Replicate 2",
                                   exp.field %in% c("20min_s8", "20min_s9") ~ "Replicate 3"), colony)) 
  

pup1.cell.attr %>%
  mutate(exp = str_split(exp.field, "_", simplify = T)[,1]) %>% split(.$exp)

$`20min`

$`40min`

$`60min`

NA

Adding the gfp.sum column for the stable GFP experiment

pup1.cell.attr <- read_csv("~/plots/pup1-rfp-gfp-decay/4-28-21-8hrGFP/data/gfp_stable_filtered.csv") %>% 
  filter(image.no == 1) %>% 
  dplyr::select(cell.id, gfpSumBgAFsub) %>% 
  mutate(cell.id = paste0(cell.id,"_","stable_pup1-rfp_none")) %>% 
  dplyr::rename("gfp.sum.bg.af.sub" = "gfpSumBgAFsub") %>% 
  left_join(pup1.cell.attr %>% 
              filter(degron == "stable") %>% 
              dplyr::select(-gfp.sum.bg.af.sub),., by = "cell.id") %>% 
  bind_rows(pup1.cell.attr %>% 
              filter(degron != "stable"),.)

removing multiple entries of the cell because some cell have more than one,

multiple.pup1 <- pup1.cell.attr %>% 
  mutate(dapi.mean.bg.sub.puncta = ifelse(is.na(dapi.mean.bg.sub.puncta), 0, dapi.mean.bg.sub.puncta) , 
         rfp.mean.bg.sub.puncta = ifelse(is.na(rfp.mean.bg.sub.puncta), 0 , rfp.mean.bg.sub.puncta)) %>% 
  filter(rfp.mean.bg.sub.puncta > 0) %>% 
  group_by(cell.id) %>% 
  tally() %>% 
  filter(n>1)

pup1.cell.attr<- pup1.cell.attr %>% 
  group_by(cell.id) %>% 
  mutate(rfp.mean.bg.sub.puncta = ifelse(cell.id %in% multiple.pup1$cell.id , mean(rfp.mean.bg.sub.puncta), rfp.mean.bg.sub.puncta),
         rfp.sum.bg.sub.puncta = ifelse(cell.id %in% multiple.pup1$cell.id, sum(rfp.mean.bg.sub.puncta), rfp.sum.bg.sub.puncta)) %>% 
    distinct(cell.id, .keep_all = TRUE) 
  

#proteasome inhibition experiments

protInhi.attr <- read_csv("~/plots/all_data/all_mg135_attr.csv") %>% 
  rename("cell.id" = "unique.trackID") %>% 
  filter(timepoint == 1) %>% 
  mutate(cell.id = paste0(cell.id ,"_",degron,"_",red,"_",treatment)) %>% 
  mutate(exp.field = paste0(str_split(cell.id, "_", simplify = T)[,2],"_",
                            str_split(cell.id, "_", simplify = T)[,3]),
    colony = case_when(exp.field %in% c("20min_s3", "20min_s4") ~ "Replicate 1", 
                            exp.field %in% c("20min_s5", "20min_s6") ~ "Replicate 2",
                            exp.field %in% c("20min_s7", "20min_s8" , "20min_s9") ~ "Replicate 3"))

multiple.proInh <- protInhi.attr %>% 
  mutate(dapi.mean.bg.sub.puncta = ifelse(is.na(dapi.mean.bg.sub.puncta), 0, dapi.mean.bg.sub.puncta) , 
         rfp.mean.bg.sub.puncta = ifelse(is.na(rfp.mean.bg.sub.puncta), 0 , rfp.mean.bg.sub.puncta)) %>% 
  filter(rfp.mean.bg.sub.puncta > 0) %>% 
  group_by(cell.id) %>% 
  tally() %>% 
  filter(n>1)

protInhi.attr <- protInhi.attr %>% 
  group_by(cell.id) %>% 
  mutate(rfp.mean.bg.sub.puncta = ifelse(cell.id %in% multiple.proInh$cell.id , mean(rfp.mean.bg.sub.puncta), rfp.mean.bg.sub.puncta),
         rfp.sum.bg.sub.puncta = ifelse(cell.id %in% multiple.proInh$cell.id, sum(rfp.mean.bg.sub.puncta), rfp.sum.bg.sub.puncta)) %>% 
    distinct(cell.id, .keep_all = TRUE)  
  

Adding the protein inhibition experiment data

pup1.proInhi.attr <- pup1.cell.attr %>% 
  bind_rows(.,protInhi.attr %>% 
              dplyr::select(-time, 
                            -timepoint,
                            -image.no, 
                            -ln.gfp, 
                            -ln.gfp.dif,
                            -trackID,
                            -gfp.intensity.center,
                            -gfp.int.mean,
                            -gfp.int.median,
                           -gfp.int.sum,
                           -no.of.triangles,
                           -field,
                           -time.dif.gfp,
                           -real.time.gfp,
                           -sample,
                           -avg.gfp.bg,
                           -Min_gfp,
                           -gfp.mean.bg.sub,
                           -value,
                           -t80,
                           -t95,
                           -threshold,
                           -threshold_95,
                           -gfp.mean.bg.af.sub,
                           -threshold_80,
                           -Mean_gfp,
                           -image,
                           -experiment,
                           -gfp.sum.bg.sub))

#keeping the parameters estimated from the 2-parameter with maturation model

#only dy.dm
dy.pup1.rep1.mat <- aic.df %>% 
  filter(red == "pup1-rfp") %>% 
  filter(ifelse(degron %in% c("stable","stable.2","stable.3"), dy < 0.1, dy < 0.5)) %>% 
  group_by(cell.id) %>% 
  filter(model == "dy.dm") %>% 
  mutate(dm = ifelse(is.na(dm), Inf, dm)) %>%
  filter( dm > 0.000015)

#count the number of cells in each experiment

dy.pup1.rep1.mat %>% group_by(degron, treatment) %>% tally()

#creating the df for the MLR

mlr.df <- dy.pup1.rep1.mat %>% 
  #left join with the df with cellular attributes
  left_join(.,pup1.proInhi.attr, by = c("cell.id","red","treatment","degron", "exp.field","colony")) %>% 
  #remove cells with no rfp puncta info 
  mutate(rfp.mean.bg.sub.puncta = ifelse(is.na(rfp.mean.bg.sub.puncta), 0, rfp.mean.bg.sub.puncta)) %>% 
  filter(rfp.mean.bg.sub.puncta > 0)

#checking how many cells there are in each experiment

mlr.df %>% group_by(degron, treatment) %>% tally()

##Removing the columns that I wont regress upon

mlr.df <- mlr.df %>% 
  filter(pos.x > 45 & pos.x < 1150,
  pos.y > 45 & pos.y < 1000) %>%
  dplyr::select(-f, 
         -rfp.mean.af.sub.puncta,
         -rfp.sum.af.sub.puncta, 
         -dapi.sum.af.sub.puncta,
         -dapi.mean.af.sub.puncta,
         -model, 
         -k,
         -fevals,
         -gevals, 
         -niter, 
         -convcode, 
         -kkt1,
         -kkt2,
         -value,
         -xtime, 
         # -treatment,
         -rfp.mean.bg.sub,
         -rfp.sum.bg.sub,
         -dapi.mean.bg.sub,
         -dapi.sum.bg.sub,
         -dapi.mean.bg.sub.puncta,
         -area.puncta,
         -BB_B,
         -BB_C,
         -pos.x,
         -pos.y,
         -gfp.sum.bg.af.sub,
         -rfp.sum.bg.sub.puncta,
         -no.of.voxels,
         -volume,
         -spheracity) %>% 
  mutate(shape = Elip_B/Elip_C, 
         t.half = log(2)/dy) %>%
  ungroup() %>%
  dplyr::select(-cell.id, 
                -red,
                -exp.field, 
                -aic,
                -Elip_B,
                -Elip_C, 
                -exp) 

#Regressors to use for MLR

regressors <- c("dm",
                "gfp.mean.bg.af.sub.new",
                # "gfp.sum.bg.af.sub",
                "area",
                # "no.of.voxels",
                # "spheracity",
                # "volume",
                "rfp.mean.bg.sub.puncta",
                # "rfp.sum.bg.sub.puncta",
                "dapi.sum.bg.sub.puncta",
                "shape")

#df where the two experiments of cln2 and stable-GFP are clubbed

mlr.df2 <- mlr.df %>% 
  filter(degron %in% c("cln2.3","mODC.2","stable.3","stable.2","cln2.4")) %>% 
  mutate(degron = case_when(degron %in% c("cln2.3","cln2.2","cln2","cln2.4") ~ "yeGFP-CLN2",
                            degron %in% c("mODC.2", "mODC")  ~ "yeGFP-mODC",
                            degron %in% c("stable","stable.2","stable.3") ~ "yeGFP")) 
  

#df to be used with step function

temp.df.list <- mlr.df2 %>% 
    filter(treatment == "none") %>%  
  # mutate( dm = log10(dm)) %>% 
  split(.$degron)

#using stepAIC from MASS

full.stepAIC.mass %>% map(.,broom::tidy)

$`GFP-CLN2`

$`GFP-mODC`

$yeGFP

NA

#step AIC with all the fp markers + cell shape

full.stepAIC <- lapply(temp.df.list, function(a){
  n <- nrow(a)
 step(lm(t.half ~ ., data = a %>% 
     dplyr::select(regressors,
                   t.half,
                   -dy, 
                   -treatment)), 
     direction = "backward", 
     trace = T,
     # k = log(n),
     k =2 ) #k = 2 when small dataset and k = log(n) where n = no. of observations for when the dataset is large
})

Note: Using an external vector in selections is ambiguous.
ℹ Use `all_of(regressors)` instead of `regressors` to silence this message.
ℹ See <https://tidyselect.r-lib.org/reference/faq-external-vector.html>.
This message is displayed once per session.
Start:  AIC=5023.53
t.half ~ dm + gfp.mean.bg.af.sub.new + area + rfp.mean.bg.sub.puncta + 
    dapi.sum.bg.sub.puncta + shape

                         Df Sum of Sq   RSS    AIC
- rfp.mean.bg.sub.puncta  1      32.7 36232 5023.0
- shape                   1      38.8 36239 5023.3
- area                    1      39.3 36239 5023.3
<none>                                36200 5023.5
- dm                      1     247.6 36447 5032.5
- dapi.sum.bg.sub.puncta  1     918.2 37118 5061.8
- gfp.mean.bg.af.sub.new  1    7805.7 44006 5335.7

Step:  AIC=5022.98
t.half ~ dm + gfp.mean.bg.af.sub.new + area + dapi.sum.bg.sub.puncta + 
    shape

                         Df Sum of Sq   RSS    AIC
- shape                   1      37.2 36270 5022.6
<none>                                36232 5023.0
- area                    1      89.4 36322 5024.9
- dm                      1     259.8 36492 5032.5
- dapi.sum.bg.sub.puncta  1     970.8 37203 5063.5
- gfp.mean.bg.af.sub.new  1    7803.3 44036 5334.8

Step:  AIC=5022.63
t.half ~ dm + gfp.mean.bg.af.sub.new + area + dapi.sum.bg.sub.puncta

                         Df Sum of Sq   RSS    AIC
<none>                                36270 5022.6
- area                    1      82.2 36352 5024.3
- dm                      1     249.7 36519 5031.7
- dapi.sum.bg.sub.puncta  1    1048.1 37318 5066.5
- gfp.mean.bg.af.sub.new  1    7848.3 44118 5335.8
Start:  AIC=1261.28
t.half ~ dm + gfp.mean.bg.af.sub.new + area + rfp.mean.bg.sub.puncta + 
    dapi.sum.bg.sub.puncta + shape

                         Df Sum of Sq    RSS    AIC
- shape                   1      1.67 3519.1 1260.1
- dm                      1      2.15 3519.6 1260.3
<none>                                3517.5 1261.3
- rfp.mean.bg.sub.puncta  1      5.47 3522.9 1261.8
- area                    1     85.84 3603.3 1299.2
- dapi.sum.bg.sub.puncta  1     89.54 3607.0 1300.9
- gfp.mean.bg.af.sub.new  1    479.97 3997.4 1471.2

Step:  AIC=1260.06
t.half ~ dm + gfp.mean.bg.af.sub.new + area + rfp.mean.bg.sub.puncta + 
    dapi.sum.bg.sub.puncta

                         Df Sum of Sq    RSS    AIC
- dm                      1      2.17 3521.3 1259.1
<none>                                3519.1 1260.1
- rfp.mean.bg.sub.puncta  1      5.05 3524.2 1260.4
- area                    1     84.64 3603.8 1297.4
- dapi.sum.bg.sub.puncta  1     87.90 3607.0 1298.9
- gfp.mean.bg.af.sub.new  1    482.37 4001.5 1470.9

Step:  AIC=1259.08
t.half ~ gfp.mean.bg.af.sub.new + area + rfp.mean.bg.sub.puncta + 
    dapi.sum.bg.sub.puncta

                         Df Sum of Sq    RSS    AIC
<none>                                3521.3 1259.1
- rfp.mean.bg.sub.puncta  1      5.13 3526.4 1259.5
- area                    1     85.97 3607.3 1297.0
- dapi.sum.bg.sub.puncta  1     87.30 3608.6 1297.7
- gfp.mean.bg.af.sub.new  1    493.27 4014.6 1474.3
Start:  AIC=922.36
t.half ~ dm + gfp.mean.bg.af.sub.new + area + rfp.mean.bg.sub.puncta + 
    dapi.sum.bg.sub.puncta + shape

                         Df Sum of Sq    RSS     AIC
<none>                                2628.7  922.36
- shape                   1      7.60 2636.3  924.18
- area                    1     49.53 2678.2  945.06
- rfp.mean.bg.sub.puncta  1     56.77 2685.5  948.63
- dapi.sum.bg.sub.puncta  1     93.01 2721.7  966.37
- dm                      1    105.69 2734.4  972.52
- gfp.mean.bg.af.sub.new  1    561.29 3190.0 1176.40

full.stepAIC %>% 
  map(., broom::glance) %>% 
  bind_rows(.id = "sample") 

  # filter(sample %in% c("mODC.2","cln2.3","cln2.4","stable.2","stable.3"))

full.stepAIC %>% 
  map(., broom::tidy) 

$yeGFP

$`yeGFP-CLN2`

$`yeGFP-mODC`

  # bind_rows(.id = "sample") %>% 
  # filter(sample %in% c("mODC.2","cln2.3","cln2.4","stable.2","stable.3"), 
         # p.value < 0.05) %>% 

full.stepAIC %>% 
  map(.,function(a){
    a$anova 
  }) %>% 
  bind_rows(.,.id = "degron") %>% 
  mutate(Step = as.character(Step), 
         Step = ifelse(Step == "", "Full",Step)) %>% 
  group_by(degron) %>% 
  mutate(del.aic = AIC[1]-AIC) 

full.stepAIC.fp %>% 
  map(.,function(a){
    a$anova 
  }) %>% 
  bind_rows(.,.id = "degron") %>% 
  mutate(Step = as.character(Step), 
         Step = ifelse(Step == "", "Full",Step)) %>% 
  group_by(degron) %>% 
  mutate(del.aic = AIC[1]-AIC) 

NA
NA

#step AIC with only the fluorescent markers

#looking at which regressor contributes to the overall variability 
full.stepAIC.fp %>% 
  map(., broom::tidy) 

$`GFP-CLN2`

$`GFP-mODC`

$yeGFP

NA

Variance in t-half explained by each regressor using anova

var.explained

$yeGFP

$`yeGFP-CLN2`

$`yeGFP-mODC`

NA

#plotting the variance explained

#plots:

gfp.6.cor.plt

`geom_smooth()` using formula 'y ~ x'

with protein inhibition

 mlr.df %>%
  mutate(dm = log10(dm)) %>% 
  filter(degron %in% c("mODC")) %>%
  # filter(!(treatment %in% c("none", "dmso1","50uM")), 
         # ifelse(treatment == "5uM", t.half < 138.6294, t.half == t.half)) %>%  
  dplyr::select(t.half,
         regressors,
         degron,
         treatment) %>%
  rename(
    "GFP" = "gfp.mean.bg.af.sub.new",
    "Pup1-tDimer" = "rfp.mean.bg.sub.puncta",
    "DNA" = "dapi.sum.bg.sub.puncta",
    "Area" = "area"
  ) %>%
  pivot_longer(cols = 2:7) %>%
  mutate(degron = case_when(
    degron %in% c("mODC")  ~ "yeGFP-mODC"),
    treatment = factor(treatment, levels = c("none","dmso1","dmso2","1uM","2.5uM","5uM","50uM"))) %>%
  filter(!(name %in% c("shape","dm")),
         !(treatment %in% c("none","dmso1","50uM")), 
         t.half < 130) %>% 
  
  ggplot(., aes(x = value, y = t.half)) +
      ggpointdensity::geom_pointdensity(size = 0.2, guides = F) +
      stat_cor(size = 2) +
      geom_smooth(method = "lm", se = FALSE, color = "red4", size = 0.5)+
      facet_grid(treatment ~ name, scales = "free") +
      theme_bw() +
      labs(x = NULL) +
      theme(text = element_text(size = 8),
            axis.text.x = element_text(angle = 30),
            legend.key.size = unit(2,"mm"), 
            legend.position = "none")+
  ylab("Half-Life [min.]")+
  xlab("A.U")

Warning: Ignoring unknown parameters: guides
`geom_smooth()` using formula 'y ~ x'

#stepAIC with proteasome inhibition experiments With all fp regressors

temp.df.list.dmso <- mlr.df %>% 
    filter(treatment != "none", t.half < 130) %>%  
  split(.$treatment)

full.stepAIC.fp.dmso <- lapply(temp.df.list.dmso, function(a){
 step(lm(t.half ~ ., data = a %>% 
     dplyr::select(regressors.fp,
                   t.half,
                   -dy, 
                   -treatment)), 
     direction = "backward", trace = T)
})

with fp regressors + celluar shape

temp.df.list.dmso <- mlr.df %>% 
    filter(treatment != "none", t.half < 130) %>%  
  split(.$treatment)

full.stepAIC.dmso <- lapply(temp.df.list.dmso, function(a){
 step(lm(t.half ~ ., data = a %>% 
     dplyr::select(regressors.fp,
                   t.half,
                   -dy, 
                   -treatment)), 
     direction = "backward", trace = T)
})

Variance in t-half explained by each regressor

full.stepAIC.dmso %>% 
  map(., anova %>% as.data.frame) %>% 
  bind_rows(.id = "sample") %>% 
  rownames_to_column(var = "predictor") %>% 
  # filter(sample %in% c("mODC.2","cln2.3","cln2.4","stable.2","stable.3")) %>% 
  group_by(sample) %>% 
  mutate(var.exp = (value.Sum.Sq/sum(value.Sum.Sq))*100) %>% 
  split(.$sample)

full.stepAIC.fp.dmso %>% 
  map(., anova %>% as.data.frame) %>% 
  bind_rows(.id = "sample") %>% 
  rownames_to_column(var = "predictor") %>% 
  # filter(sample %in% c("mODC.2","cln2.3","cln2.4","stable.2","stable.3")) %>% 
  group_by(sample) %>% 
  mutate(var.exp = (value.Sum.Sq/sum(value.Sum.Sq))*100) %>% 
  split(.$sample)

full.stepAIC.dmso %>% 
  map(., anova %>% as.data.frame) %>% 
  bind_rows(.id = "sample") %>% 
  rownames_to_column(var = "predictor") %>% 
  # filter(sample %in% c("mODC.2","cln2.3","cln2.4","stable.2","stable.3")) %>% 
  group_by(sample) %>% 
  mutate(var.exp = (value.Sum.Sq/sum(value.Sum.Sq))*100, 
         predictor = str_split(predictor, pattern = "\\.\\.\\.", simplify = T)[,1], 
         predictor = case_when(predictor == "gfp.mean.bg.af.sub.new" ~ "GFP" ,
                               predictor == "area" ~ "Area",
                               predictor == "rfp.mean.bg.sub.puncta" ~ "Pup1",
                               predictor == "dapi.sum.bg.sub.puncta" ~ "DNA" ,
                               predictor == "shape" ~ "Shape", 
                               predictor == "dm" ~ "Rate of Maturation"
                               , 
                               TRUE~predictor)) %>% 
  split(.$sample)

$`1uM`

$`2.5uM`

$`50uM`

$`5uM`

$dmso1

$dmso2

NA

mlr.df %>%
  mutate(dm = log10(dm)) %>% 
  filter(degron %in% c ("mODC") ) %>%
  filter(!(treatment %in% c("none", "dmso1","50uM")), 
         ifelse(treatment == "5uM", t.half < 138.6294, t.half == t.half)) %>% 
  dplyr::select(t.half,
         regressors,
         degron,
         treatment) %>% 
  ggplot(.,aes(x = area, y = rfp.mean.bg.sub.puncta))+
  geom_point()+
  geom_smooth(method = "lm", se = F)+
  facet_wrap(~treatment)+
  stat_cor()

#ggpairs

gfp.area.cor.ggPair

$`yeGFP-mODC`

 plot: [1,1] [====>----------------------------------------------------------------------------------]  6% est: 0s 
 plot: [1,2] [==========>----------------------------------------------------------------------------] 12% est: 1s 
 plot: [1,3] [===============>-----------------------------------------------------------------------] 19% est: 1s 
 plot: [1,4] [=====================>-----------------------------------------------------------------] 25% est: 1s 
 plot: [2,1] [==========================>------------------------------------------------------------] 31% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [2,2] [================================>------------------------------------------------------] 38% est: 1s 
 plot: [2,3] [=====================================>-------------------------------------------------] 44% est: 1s 
 plot: [2,4] [===========================================>-------------------------------------------] 50% est: 1s 
 plot: [3,1] [================================================>--------------------------------------] 56% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [3,2] [=====================================================>---------------------------------] 62% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [3,3] [===========================================================>---------------------------] 69% est: 0s 
 plot: [3,4] [================================================================>----------------------] 75% est: 0s 
 plot: [4,1] [======================================================================>----------------] 81% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,2] [===========================================================================>-----------] 88% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,3] [=================================================================================>-----] 94% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,4] [=======================================================================================]100% est: 0s 
                                                                                                                   

$`yeGFP-CLN2`

 plot: [1,1] [====>----------------------------------------------------------------------------------]  6% est: 0s 
 plot: [1,2] [==========>----------------------------------------------------------------------------] 12% est: 0s 
 plot: [1,3] [===============>-----------------------------------------------------------------------] 19% est: 1s 
 plot: [1,4] [=====================>-----------------------------------------------------------------] 25% est: 1s 
 plot: [2,1] [==========================>------------------------------------------------------------] 31% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [2,2] [================================>------------------------------------------------------] 38% est: 1s 
 plot: [2,3] [=====================================>-------------------------------------------------] 44% est: 1s 
 plot: [2,4] [===========================================>-------------------------------------------] 50% est: 1s 
 plot: [3,1] [================================================>--------------------------------------] 56% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [3,2] [=====================================================>---------------------------------] 62% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [3,3] [===========================================================>---------------------------] 69% est: 0s 
 plot: [3,4] [================================================================>----------------------] 75% est: 0s 
 plot: [4,1] [======================================================================>----------------] 81% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,2] [===========================================================================>-----------] 88% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,3] [=================================================================================>-----] 94% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,4] [=======================================================================================]100% est: 0s 
                                                                                                                   

$yeGFP

 plot: [1,1] [====>----------------------------------------------------------------------------------]  6% est: 0s 
 plot: [1,2] [==========>----------------------------------------------------------------------------] 12% est: 0s 
 plot: [1,3] [===============>-----------------------------------------------------------------------] 19% est: 1s 
 plot: [1,4] [=====================>-----------------------------------------------------------------] 25% est: 1s 
 plot: [2,1] [==========================>------------------------------------------------------------] 31% est: 1s `geom_smooth()` using formula 'y ~ x'

 plot: [2,2] [================================>------------------------------------------------------] 38% est: 1s 
 plot: [2,3] [=====================================>-------------------------------------------------] 44% est: 0s 
 plot: [2,4] [===========================================>-------------------------------------------] 50% est: 0s 
 plot: [3,1] [================================================>--------------------------------------] 56% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [3,2] [=====================================================>---------------------------------] 62% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [3,3] [===========================================================>---------------------------] 69% est: 0s 
 plot: [3,4] [================================================================>----------------------] 75% est: 0s 
 plot: [4,1] [======================================================================>----------------] 81% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,2] [===========================================================================>-----------] 88% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,3] [=================================================================================>-----] 94% est: 0s `geom_smooth()` using formula 'y ~ x'

 plot: [4,4] [=======================================================================================]100% est: 0s 
                                                                                                                   

ggsave(plot = gfp.area.cor.ggPair$`yeGFP-mODC`, filename = "gfp_mODC_4Cor_ggP.png", path = "~/plots/paper1/figures/fig_2/", width = 4, height = 3.5)

ggsave(plot = gfp.area.cor.ggPair$`yeGFP-mODC`, filename = "gfp_mODC_4Cor_ggP.pdf", path = "~/plots/paper1/figures/fig_2/", width = 4, height = 3.5)

ggsave(plot = gfp.area.cor.ggPair$`yeGFP-CLN2`, filename = "gfp_cln2_4Cor_ggP.png", path = "~/plots/paper1/figures/fig_2/", width = 4, height = 3.5)

ggsave(plot = gfp.area.cor.ggPair$`yeGFP-CLN2`, filename = "gfp_cln2_4Cor_ggP.pdf", path = "~/plots/paper1/figures/fig_2/", width = 4, height = 3.5)

ggsave(plot = gfp.area.cor.ggPair$`yeGFP`, filename = "gfp_4Cor_ggP.png", path = "~/plots/paper1/figures/fig_2/", width = 4, height = 3.5)

ggsave(plot = gfp.area.cor.ggPair$`yeGFP`, filename = "gfp_4Cor_ggP.pdf", path = "~/plots/paper1/figures/fig_2/", width = 4, height = 3.5)

test.ggP <- (wrap_elements(ggmatrix_gtable(gfp.area.cor.ggPair$`yeGFP-mODC`))|wrap_elements(ggmatrix_gtable(gfp.area.cor.ggPair$`yeGFP-CLN2`))| wrap_elements(ggmatrix_gtable(gfp.area.cor.ggPair$yeGFP)))+
 plot_annotation(tag_levels = "A")

ggsave(plot = test.ggP, filename = "test_ggP.pdf", path = "~/plots/paper1/figures/fig_2/", width = 6.5, height = 3)

#area vs pup1 vs dna vs gfp

df_corSup_fig <- mlr.df %>%
  filter(degron %in% c("mODC.2", "cln2.3", "cln2.4", "stable.2", "stable.3")) %>%
  dplyr::select(t.half,
         regressors,
         degron) %>%
  rename(
    "GFP" = "gfp.mean.bg.af.sub.new",
    "Pup1-tDimer" = "rfp.mean.bg.sub.puncta",
    "DNA" = "dapi.sum.bg.sub.puncta",
    "Area" = "area"
  ) %>%
  dplyr::select(Area, DNA, `Pup1-tDimer`,GFP,degron ) %>% 
  mutate(
    degron = case_when(
      degron %in% c("cln2.3", "cln2.4") ~ "yeGFP-CLN2",
      degron %in% c("mODC.2")  ~ "yeGFP-mODC",
      degron %in% c("stable.2", "stable.3") ~ "yeGFP"
    ),
    degron = factor(degron , levels = c("yeGFP-mODC", "yeGFP-CLN2", "yeGFP"))
  ) 

area.pup1 <- df_corSup_fig %>% 
  filter(degron == "yeGFP-mODC") %>% 
  ggplot(.,aes(x = Area, y = `Pup1-tDimer`))+
  ggpointdensity::geom_pointdensity(size = 0.2, guides = F) +
      stat_cor(size = 3) +
      geom_smooth(method = "lm", se = FALSE, color = "red4", size = 0.5)+
      theme_bw() +
      labs(x = NULL) +
      theme(text = element_text(size = 8),
            axis.text.x = element_text(angle = 30),
            legend.key.size = unit(2,"mm"), 
            legend.position = "none", 
            axis.line = element_line(size = 0.1))+
  ylab("Pup1-tDimer [A.U]")+
  xlab("Area [A.U]")

Warning: Ignoring unknown parameters: guides

dna.pup1 <- df_corSup_fig %>% 
  filter(degron == "yeGFP-mODC") %>% 
  ggplot(.,aes(x = DNA, y = `Pup1-tDimer`))+
  ggpointdensity::geom_pointdensity(size = 0.2, guides = F) +
      stat_cor(size = 3) +
      geom_smooth(method = "lm", se = FALSE, color = "red4", size = 0.5)+
      theme_bw() +
      labs(x = NULL) +
      theme(text = element_text(size = 8),
            axis.text.x = element_text(angle = 30),
            legend.key.size = unit(2,"mm"), 
            legend.position = "none", 
            axis.line = element_line(size = 0.1))+
  ylab("Pup1-tDimer [A.U]")+
  xlab("DNA [A.U]")

Warning: Ignoring unknown parameters: guides

dna.area <- df_corSup_fig %>%
  filter(degron == "yeGFP-mODC") %>% 
  ggplot(.,aes(x = DNA, y = Area))+
  ggpointdensity::geom_pointdensity(size = 0.2, guides = F) +
      stat_cor(size = 3) +
      geom_smooth(method = "lm", se = FALSE, color = "red4", size = 0.5)+
      theme_bw() +
      labs(x = NULL) +
      theme(text = element_text(size = 8),
            axis.text.x = element_text(angle = 30),
            legend.key.size = unit(2,"mm"), 
            legend.position = "none", 
            axis.line = element_line(size = 0.1))+
  ylab("Area [A.U]")+
  xlab("DNA [A.U]")

Warning: Ignoring unknown parameters: guides

gfpInt.cor.plt <- df_corSup_fig %>% 
  pivot_longer(cols = 1:3) %>% 
  # filter(degron == "yeGFP-mODC") %>% 
  ggplot(.,aes(x = value, y = GFP))+
  ggpointdensity::geom_pointdensity(size = 0.2, guides = F) +
      stat_cor(size = 3) +
      geom_smooth(method = "lm", se = FALSE, color = "red4", size = 0.5)+
  facet_wrap(degron~name, scales = "free")+
      theme_bw() +
      labs(x = NULL) +
      theme(text = element_text(size = 8),
            axis.text.x = element_text(angle = 30),
            legend.key.size = unit(2,"mm"), 
            legend.position = "none", 
            axis.line = element_line(size = 0.1))+
  xlab("A.U")

Warning: Ignoring unknown parameters: guides

test_cor

`geom_smooth()` using formula 'y ~ x'
`geom_smooth()` using formula 'y ~ x'
`geom_smooth()` using formula 'y ~ x'
`geom_smooth()` using formula 'y ~ x'

#Partial correlations ppcor

#partial correlations
ppCor.list <- mlr.df2 %>% 
  filter(treatment == "none") %>% split(.$degron) %>% 
  map(.,function(a){
    a <- a %>% 
      dplyr::select(t.half, area,  dapi.sum.bg.sub.puncta, rfp.mean.bg.sub.puncta, gfp.mean.bg.af.sub.new) %>% 
      na.omit()
    ppcor::pcor(a)
  })

#correlations
cor.all <- mlr.df2 %>% 
  filter(treatment == "none") %>%  
  split(.$degron) %>% 
  map(.,function(a){
     a <- a %>% 
      dplyr::select(t.half, area, dapi.sum.bg.sub.puncta,rfp.mean.bg.sub.puncta, gfp.mean.bg.af.sub.new) %>% 
      na.omit()
     cor(a)
  })

ppCor.df <- ppCor.list %>% 
  map(.,function(a){
    a$estimate %>% 
      as.data.frame() %>% 
      rownames_to_column(var = "regressor") %>% 
      dplyr::select(t.half, regressor) %>% 
      left_join(.,a$p.value%>% 
                  as.data.frame() %>% 
                  rownames_to_column(var = "regressor") %>% 
                  rename("p.value" = "t.half") %>% 
      dplyr::select(p.value, regressor), by = "regressor")
  }) %>% 
  bind_rows(.id = "degron")

single.core.df <- cor.all %>% 
  map(.,function(a){
    a%>% 
      as.data.frame() %>% 
      rownames_to_column(var = "regressor") %>% 
      dplyr::select(t.half, regressor)
  }) %>% 
  bind_rows(.id = "degron") %>% 
  rename("single.core"="t.half")

ppCoor.fig <- ppCor.df %>% 
  left_join(.,single.core.df, by = c("degron","regressor")) %>% 
  rename("Partial-Pariwise"="t.half",
         "Standard-Pariwise" = "single.core") %>% 
  pivot_longer(cols = c("Partial-Pariwise","Standard-Pariwise")) %>% 
   mutate(regressor = case_when(regressor == "gfp.mean.bg.af.sub.new" ~ "GFP" ,
                               regressor == "area" ~ "Area",
                               regressor == "rfp.mean.bg.sub.puncta" ~ "Pup1",
                               regressor == "dapi.sum.bg.sub.puncta" ~ "DNA" ,
                               regressor == "t.half" ~ "half-life"), 
          degron = factor(degron, levels = c("yeGFP-mODC","yeGFP-CLN2","yeGFP")), 
          name = factor(name , levels = c("Standard-Pariwise", "Partial-Pariwise")), 
          sig = stars.pval(p.value), 
          sig = ifelse(sig == " ", "NS", sig)) %>% 
  # filter(p.value < 0.05) %>% 
  filter(regressor != "half-life") %>% 
  ggplot(.,aes(x = regressor , y = value, fill = name))+
  geom_col(width = 0.5, position = "dodge")+
  facet_wrap(~degron)+
  geom_hline(yintercept = 0, size = 0.1)+
  scale_y_continuous(breaks = seq(-0.5,0.5, by = 0.1))+
  theme_pubr()+
  theme(text = element_text(size = 8),
            legend.key.size = unit(4,"mm"),
        legend.position = "top", 
        axis.line = element_line(size = 0.1),
        strip.background = element_blank())+
  scale_fill_manual(values = c("lightsteelblue4" ," steelblue4"), name = NULL)+
  ylab("Correlation with Half-life ")+
  xlab("Cellular features")+
  labs(title = "C")

ppCoor.fig

#patchwork for 4 cellular features (area, gfp, dapi, pup1) and t-half

gfp.4cor.hf.ppCor.plt

`geom_smooth()` using formula 'y ~ x'

ggsave(plot = gfp.4cor.hf.ppCor.plt, filename = "gfp_4cor_halfLife_ppcor.png", path = "~/plots/paper1/figures/fig_2/", width = 7, height = 9)

`geom_smooth()` using formula 'y ~ x'

#with protein inhibion

ppCor.PrtIn.list <- mlr.df %>% 
    filter(treatment != "none") %>%  
  filter(ifelse(treatment == "5uM", t.half < 138.6294, t.half == t.half)) %>% 
  split(.$treatment) %>% 
  map(.,function(a){
    a <- a %>% 
      dplyr::select(t.half, area, dapi.sum.bg.sub.puncta, rfp.mean.bg.sub.puncta, gfp.mean.bg.af.sub.new) %>%
      na.omit()
    ppcor::pcor(a)
  })

cor.PrtIn.all <- mlr.df %>% 
    filter(treatment != "none") %>%  
  filter(ifelse(treatment == "5uM", t.half < 138.6294, t.half == t.half)) %>% 
  split(.$treatment) %>% 
  map(.,function(a){
    a <- a %>% 
      dplyr::select(t.half, area, dapi.sum.bg.sub.puncta, rfp.mean.bg.sub.puncta, gfp.mean.bg.af.sub.new) %>% 
      na.omit()
    cor(a)
  })

ppCor.PrtIn.df <- ppCor.PrtIn.list %>% 
  map(.,function(a){
    a$estimate %>% 
      as.data.frame() %>% 
      rownames_to_column(var = "regressor") %>% 
      dplyr::select(t.half, regressor) %>% 
      left_join(.,a$p.value%>% 
                  as.data.frame() %>% 
                  rownames_to_column(var = "regressor") %>% 
                  rename("p.value" = "t.half") %>% 
      dplyr::select(p.value, regressor), by = "regressor")
  }) %>% 
  bind_rows(.id = "treatment")

single.core.Prtn.df <- cor.PrtIn.all %>% 
  map(.,function(a){
    a%>% 
      as.data.frame() %>% 
      rownames_to_column(var = "regressor") %>% 
      dplyr::select(t.half, regressor)
  }) %>% 
  bind_rows(.id = "treatment") %>% 
  rename("single.core"="t.half")

correlation of pup1 with mODC with treated with proteasome inhibitors

#when you normalize pup1 with the area

mlr.df %>%
  mutate(dm = log10(dm)) %>% 
  filter(degron %in% c("mODC.2", "cln2.3", "cln2.4", "stable.2", "stable.3")) %>%
  dplyr::select(t.half,
         regressors,
         degron) %>%
  rename(
    "GFP" = "gfp.mean.bg.af.sub.new",
    "Pup1-tDimer" = "rfp.mean.bg.sub.puncta",
    "DNA" = "dapi.sum.bg.sub.puncta",
    "Area" = "area"
  ) %>%
  mutate(norm.pup1 = `Pup1-tDimer`/Area) %>% 
  pivot_longer(cols = c(2:7,9)) %>%
  mutate(degron = case_when(
      degron %in% c("cln2.3", "cln2.2", "cln2", "cln2.4") ~ "yeGFP-CLN2",
      degron %in% c("mODC.2", "mODC")  ~ "yeGFP-mODC",
      degron %in% c("stable", "stable.2", "stable.3") ~ "yeGFP"
    ),
    degron = factor(degron , levels = c("yeGFP-mODC", "yeGFP-CLN2", "yeGFP"))
  ) %>%
  filter(!(name %in% c("shape","dm"))) %>% 
  ggplot(., aes(x = value, y = t.half)) +
      ggpointdensity::geom_pointdensity(size = 0.2, guides = F) +
      stat_cor(size = 2) +
      geom_smooth(method = "lm", se = FALSE, color = "red4", size = 0.5)+
      facet_wrap(degron ~ name, scales = "free", ncol = 4) +
      theme_bw() +
      labs(x = NULL) +
      theme(text = element_text(size = 8),
            axis.text.x = element_text(angle = 30),
            legend.key.size = unit(2,"mm"), 
            legend.position = "none")+
  ylab("Half-Life [min.]")+
  xlab("A.U")